Auto Focus Area Detection Using Depth Information

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing imaging apparatuses with multiple AF areas struggle to focus on the desired object when there are multiple objects at different distances within an AF area, often focusing on an undesired object or an intermediate position, rendering such AF areas ineffective for auto focusing.

Innovation Solution

A detection apparatus and method that identifies an effective AF area by detecting the object area within an image and computing depth information from defocus data to determine focus areas with minimal distance variations, thereby excluding areas with multiple objects at different distances.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Area of stationary object

If multiple AF areas are used for auto focusing, then the coverage area for focusing is increased, but the focusing accuracy deteriorates when multiple objects at different distances exist within an AF area

Engineering Contradiction:

ImproveAF area coverageVSAvoidfocusing accuracy

Core Design Contradiction:

Area of stationary objectVSMeasurement precision

Solution Approach 1:

The image is divided into multiple focus areas, and each focus area is further segmented into multiple object areas through object detection. This hierarchical segmentation allows the system to evaluate focusing effectiveness at a finer granularity, selecting only those object areas within focus areas that contain single-distance objects for auto focusing operations.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent applies local quality by evaluating and selecting focus areas based on their specific characteristics (presence of multiple objects at different distances). Each focus area is assessed individually using object detection results, and only locally suitable areas (those without multiple distance objects) are selected for auto focusing, rather than uniformly applying focusing across all areas.

Inventive Principle:

Principle #3Local quality

2Measurement precision

If object detection is performed to identify effective AF areas, then focusing accuracy is improved, but the processing time and computational load increase

Engineering Contradiction:

Improveeffective AF area detection accuracyVSAvoidprocessing time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

Object detection and effective focus area identification are performed as preliminary actions before the actual auto focusing operation. By pre-identifying and selecting only effective focus areas that contain single-distance objects, the system avoids unnecessary focusing computations on ineffective areas, thereby reducing overall processing time despite the added initial detection step.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system performs object detection across the entire image to identify all object areas, but then selectively applies auto focusing only to the subset of focus areas determined to be effective. This partial action approach avoids the excessive computation that would result from processing all focus areas uniformly, optimizing the balance between detection thoroughness and processing efficiency.

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentUS11275230B2Detection apparatus, detection method, detection program, and imaging apparatus

Publication Date: 2022.03.15 SONY GROUP CORP

AI summary

A detection apparatus has a detector for detecting, on the basis of an object area that is an area corresponding to an object in an image, an effective focus area from two or more focus areas set to the image.